Your data warehouse
is a different animal than most of your other cloud systems, due to:
The number of system interconnections
. From the front side and all the systems where you perform Extract, Transform, and Load (ETL)
workloads, to the back side and all the different ways that the data is consumed, the number of system interconnects can easily cross into the hundreds, sometimes thousands. Even Internet of Things (IoT) systems with their large number of connections don’t handle that much data per connection.
The sheer amount of data
. Not many operational systems deal in terabytes
of data. The simple act of copying data at this scale is never instantaneous and can take longer than you think. The typical response to this issue is to throw more (very expensive) bandwidth at the problem.
When we talk to customers considering cloud implementations, they are often concerned about vendor lock-in. While vendor lock-in can be a problem, it is usually not insurmountable.The major cloud vendors tend to have very similar capabilities when it comes to underlying infrastructure and applications. Often the applications used on one vendor can be installed and used on another with little or no change. With Teradata Everywhere, we support the two big cloud vendors, Microsoft Azure
and Amazon Web Services
. Teradata used on either platform (or on-premises or with VMware) operate identically to each other. You can migrate your applications just by repointing the application with no recoding necessary.
So, what are the real issues with switching cloud providers?
- Not having a good business reason and not knowing why you are migrating. This hiccup happens more often than most people think. We see it mostly when IT departments are chasing the “shiny new thing.” Without a clear vision of why you are changing providers, you don’t know when you are finished. Are you just trading headaches?
- Data egress charges that sneak up on you. We have seen customers moving 10-20 terabytes a day between cloud providers and then getting a shocking bill at the end of the month. Unlike operational systems that deal in small bursts of data, analytic platforms routinely churn through very large amounts of information. Costs add up quickly.
- While cloud functionality is typically matched up, it’s normally accomplished in slightly different ways. Consider Azure Resource Groups versus AWS Virtual Private Clouds (VPCs). Both are used for grouping but accomplish it in different ways that don’t directly translate. You can spend a lot of time and money trying to mimic one cloud functionality in another and while not adopting the provider’s native model.
- Lastly, but by no means least, don’t underestimate educating your IT user community. While we like to talk about how it’s “transparent to the user,” large migrations rarely ever are. Have you discussed a cloud change with your internal users and what it means to them?
There are other reasons why it can be difficult to move from one cloud provider to another but working with Teradata won’t be one of them. Teradata IntelliCloud
, our as-a-service offering for analytics at scale, can serve as the core of your analytic ecosystem and make this easier for you. We remove the non-productive (but required) work from your to-do list, such as system administration and maintenance, so that you can focus on analytics and insight.
To learn more, follow the conversation at #CloudExperts or #BuiltForTheCloud, or reach out to your Teradata account team.
Ron supports international architecture and key needs for Teradata’s cloud initiative. With 25 years’ experience (including 15+ years using Teradata) in enterprise data management and strategy, Ron has diverse experience with Fortune 200 leaders in the agriculture, finance, telecommunications, health care, government and consumer goods sectors. He is involved in conceiving, creating and bringing to fruition innovations; combining thinking big with laboring through the pains of materializing ideas.
View all posts by Ron Luebke